System and method for image processing of medical test results using generalized curve field transform
A method for image processing medical self-test results receives a digital image of a visual indication of a test result. A digital image is generated of the visual indication of the test result that includes noise and distortions therein. The digital image is processed using generalized curve field transforms to extract relevant features of the digital image in a presence of the noise and distortions to create a transformed image. A diagnosis is generated based upon the transformed image to the plurality of images of the test results.
1. A method for image processing medical self-test results, comprising:
receiving a self-test result that provides a visual indication of a test result;
generating a digital image of the visual indication of the test result, the digital image including noise and distortions therein;
processing the digital image using generalized curve field transforms to extract relevant geometrical features of the digital image that are not distorted by the noise and distortions within the digital image to create a transformed image that is not distorted by the noise and distortions of the digital image;
generating a diagnosis based upon the transformed image that is not distorted by the noise and distortions of the digital image; and
displaying the diagnosis.
2. The method of claim 1 , wherein the step of processing further comprises:
generating an alignment spectrum image of the digital image;
generating a generalized curve field transformation of the digital image from the alignment spectrum image; and
determining a presence of at least one geometric shape within the alignment spectrum image.
3. The method of claim 2 further comprising the step of filtering the alignment spectrum image to remove at least one predetermined geometric shape.
4. The method of claim 3 further including the step of determining a presence of a second at least one predetermined geometric shape within the alignment spectrum image.
5. The method of claim 2 , wherein the step of generating an orientation field further comprises:
determining for each pixel within the digital image an orientation line from a plurality of possible lines based upon an intensity response for each orientation line; and
determining for each pixel within the digital image a reliability factor for the determined orientation line.
6. The method of claim 1 , wherein the step of processing further comprises removing shadows from the digital image to create the transformed image that is not distorted by the noise and distortions of the digital image.
7. A method for image processing medical self-test results, comprising:
receiving a digital image of a visual indication of a test result, the digital image including noise and distortions therein;
processing the digital image using generalized curve field transforms to extract relevant geometrical features of the digital image that are not distorted by the noise and distortions within the digital image to create a transformed image that is not distorted by the noise and distortions of the digital image; and
generating a diagnosis based upon the transformed image that is not distorted by the noise and distortions of the digital image.
8. The method of claim 7 , wherein the step of processing further comprises:
generating an alignment spectrum of the digital image;
generating a generalized curve field transformation of the digital image from the alignment spectrum image; and
determining a presence of at least one geometric shape within the alignment spectrum image.
9. The method of claim 8 further comprising the step of filtering the alignment spectrum image to remove at least one predetermined geometric shape.
10. The method of claim 9 further including the step of determining a presence of a second at least one predetermined geometric shape within the alignment spectrum image.
11. The method of claim 8 , wherein the step of generating an orientation field further comprises:
determining for each pixel within the digital image an orientation line from a plurality of possible orientation lines based upon an intensity response for the orientation line; and
determining for each pixel within the digital image a reliability factor for the determined orientation line.
12. The method of claim 7 , wherein the step of processing further comprises removing shadows from the digital image to create the transformed image that is not distorted by the noise and distortions of the digital image.
13. A method for image processing medical self-test results, comprising:
receiving a self-test result that provides a visual indication of a test result;
generating a digital image of the visual indication of the test result, the digital image including noise and distortions therein;
processing the digital image using generalized curve field transforms to extract relevant geometrical features of the digital image that are not distorted by the noise and distortions within the digital image to create a transformed image that is not distorted by the noise and distortions of the digital image, the step of processing further comprising:
generating an alignment spectrum of the digital image;
generating a generalized curve field transformation of the digital image from the alignment spectrum;
determining a presence of at least one geometric shape within the alignment spectrum image;
determining a presence of a second at least one predetermined geometric shape within the alignment spectrum image;
filtering the alignment spectrum image to remove the second at least one predetermined geometric shape;
generating the transformed image that is not distorted by the noise and distortions of the digital image responsive to the generalized curve field transform and the filtered alignment spectrum image;
generating a diagnosis based upon the transformed image that is not distorted by the noise and distortions of the digital image; and
displaying the diagnosis.
14. The method of claim 13 , wherein the step of generating an orientation field further comprises:
determining for each pixel within the digital image an orientation line from a plurality of possible orientation lines based upon an intensity response for the orientation line; and
determining for each pixel within the digital image a reliability factor for the determined orientation line.
15. The method of claim 13 , wherein the step of processing further comprises removing shadows from the digital image to create the transformed image that is not distorted by the noise and distortions of the digital image.